Metadata Amazon didn't work

[redskull1986]

What went wrong?

If I fetch metadata with Amazon.de, I didn't get the full metadata.

In detail, every time Grimmory fetch metadata and find a physical book on Amzon.de I get full metadata:

Most time (and this is the preferred format) Grimmory fetch metadata from kindle format at Amazon.de an I didn't get the metadata

Also in both matches no series numbers:

On the first look into the code ( AmazonBookParser.java ) and the site (amazon.de) it search for the element

averageCustomerReviews_feature_div

private Integer getReviewCount(Document doc) {
    try {
        Element reviewDiv = doc.select("div#averageCustomerReviews_feature_div").first();
        if (reviewDiv != null) {
            Element reviewCountElement = reviewDiv.getElementById("acrCustomerReviewText");
            if (reviewCountElement != null) {
                String reviewCountRaw = reviewCountElement.text().split(" ")[0];
                String reviewCountClean = NON_DIGIT_PATTERN.matcher(reviewCountRaw).replaceAll("");
                if (!reviewCountClean.isEmpty()) {
                    return Integer.parseInt(reviewCountClean);
                }
            }
        }
    } catch (Exception e) {
        log.warn("Failed to parse amazonReviewCount: {}", e.getMessage());
    }
    return null;
}

In the pysical book site, the element exist:

id="averageCustomerReviews_feature_div" class="celwidget" data-feature-name="averageCustomerReviews"

In the kindle format site, the elements name is different:

id="reviewFeatureGroup" class="celwidget" data-feature-name="reviewFeatureGroup"

The ratings are the same element.

The pages are:

rpi-attribute-book_details-fiona_pages

and

rpi-attribute-book_details-ebook_pages

and so on.

How can we reproduce it?

Fetch metadata from amazon(.de) first from the physical and second from the kindle version.

What Build of Grimmory are you on?

Stable

Your setup

• Version 3.0.2 (2.9.x same bug)
• Docker installation
• Debian

Screenshots or error messages (optional)

No response

Before submitting

• I've searched existing discussions and this hasn't been reported yet
• I've reviewed this report and have completed it as fully as I can

[redskull1986]

I've looked on the code and have a working parser now:

``

Series numbers, now for German, English and France:

private static final Pattern SERIES_FORMAT_PATTERN = Pattern.compile("(?:Buch|Book|Livre) (\\d+(?:\\.\\d+)?) (?:von|of|sur) (\\d+)");

Rating:

private Double getRating(Document doc) {
    try {
        Element reviewDiv = doc.select("div#reviewFeatureGroup").first();
        if (reviewDiv == null) {
            reviewDiv = doc.select("div#averageCustomerReviews_feature_div").first();
        }
        if (reviewDiv != null) {
            Element ratingSpan = reviewDiv.selectFirst("span#acrPopover span.a-size-base.a-color-base");
            if (ratingSpan == null) {
                ratingSpan = reviewDiv.selectFirst("span#acrPopover span.a-size-small.a-color-base");
            }
            if (ratingSpan != null) {
                String text = ratingSpan.text().trim();
                if (!text.isEmpty()) {
                    String normalizedText = text.replace(',', '.');
                    return Double.parseDouble(normalizedText);
                }
            }
        }
    } catch (Exception e) {
        log.warn("Failed to parse amazonRating: {}", e.getMessage());
    }
    return null;
}

Reviews:

private Integer getReviewCount(Document doc) {
    try {
        Element reviewDiv = doc.select("div#reviewFeatureGroup").first();
        if (reviewDiv == null) {
            reviewDiv = doc.select("div#averageCustomerReviews_feature_div").first();
        }
        if (reviewDiv != null) {
            Element reviewCountElement = reviewDiv.getElementById("acrCustomerReviewText");
            if (reviewCountElement != null) {
                String reviewCountRaw = reviewCountElement.text().split(" ")[0];
                String reviewCountClean = NON_DIGIT_PATTERN.matcher(reviewCountRaw).replaceAll("");
                if (!reviewCountClean.isEmpty()) {
                    return Integer.parseInt(reviewCountClean);
                }
            }
        }
    } catch (Exception e) {
        log.warn("Failed to parse amazonReviewCount: {}", e.getMessage());
    }
    return null;
}

Pages:

private Integer getPageCount(Document doc) {
    Elements pageCountElements = doc.select("#rpi-attribute-book_details-ebook_pages .rpi-attribute-value span");
    if (pageCountElements.isEmpty()) {
        pageCountElements = doc.select("#rpi-attribute-book_details-fiona_pages .rpi-attribute-value span");
    }
    if (!pageCountElements.isEmpty()) {
        String pageCountText = pageCountElements.first().text();
        if (!pageCountText.isEmpty()) {
            try {
                String cleanedPageCount = NON_DIGIT_PATTERN.matcher(pageCountText).replaceAll("");
                return Integer.parseInt(cleanedPageCount);
            } catch (NumberFormatException e) {
                log.warn("Error parsing page count: {}, error: {}", pageCountText, e.getMessage());
            }
        }
    }
    return null;
}

This are only minor changes to get it work again. I hope it would be implementet soon.

[imnotjames]

Can you include links to the books on Amazon you see this exhibited on?

Does the kindle format site issue apply to other Amazon locales?

[imnotjames]

Also can you open a separate bug report specific to the series name problem. It's a separate issue from the kindle format site.

[redskull1986]

You can find it here #1128

[redskull1986]

Here are some links

German:
Ebook
Hardcover

France:
Ebook
Hardcover

The issue is the same on the france site, even with german and france books.

[redskull1986]

Do you need any additional information to implement this?